E L S E V I E R Brain Research 649 (1994) 208 216

BRAIN RESEARCH

Research Report

Evidence for functionally important adenosine Aza receptors in the rat hippocampus

Rodrigo A. Cunha a,b,* Bj6rn Johansson b Ingeborg van der Ploeg b, Ana M. Sebastifio ~ J. Alexandre Ribeiro aT Bertil B. Fredholrn b

Laboratory of Pharmacology, Gulbenkian Institute of Science, 2781 Oeiras codex, Portugal I, Section of Molecular Neuropharmacology, Department of Physiology and Pharmacology, Karolinska lnstitutet, S-171 77 Stockhohn, Sweden

(Accepted 15 March 1994)

Abstract

Adenosine A2a receptors are not confined to dopamine-rich areas of the brain, since thermocycling analysis shows that adenosine A2a receptor mRNA is expressed also in the hippocampus (CA1, CA3 and dentate gyrus) and cerebral cortex. The expression of Aza mRNA in three main areas of the hippocampus was confirmed by in situ hybridization; A2~ mRNA expression was mainly localized in the pyramidal and granular cells, the same hippocampal regions that showed adenosine A I receptor mRNA expression. Receptor autoradiographic studies with [3H]CGS 21680 (30 nM), a selective adenosine Az~ receptor agonist, showed specific binding sites in the hippocampus. The density of [3H]CGS 21680 binding was greatest in the stratum radiatum of the CA1 area, followed by the stratum oriens of the cornu Ammonis, stratum radiatum of the CA3 area and supra-granular layer of the dentate gyrus. This anatomical distribution of [3H]CGS 21680 binding was similar to the pattern of [3H]CHA binding in the hippocampus. Electrophysiological studies in the Schaffer fibers/CA1 pyramids showed that upon activation of the A2~ , receptors with CGS 21680 (10 nM) the ability of the adenosine A~ receptor agonist, CPA, to inhibit neuronal activity was significantly attenuated. These results show functionally important co-expression and co-localization of adenosine A2~, and A l receptors in the hippocampus. The results also suggest that adenosine A2a receptor-mediated neuromodulation is not confined to the basal ganglia, but is more widespread throughout the nervous system.

Key words: Adenosine; A2a receptor; A 1 receptor; Hippocampus

I. Introduct ion

Adenosine is a neuromodulator which generally de- presses neuronal excitability in the central and periph- eral nervous system via an action on pre- and postsy- naptic adenosine A 1 receptors [4]. A 1 receptors are present in high amounts in the hippocampal formation [6]. There is also evidence that adenosine can stimulate neurotransmitter release via an action on adenosine A2a receptors in the striatum and innervated skeletal muscle [1,2]. Adenosine Aza receptors in the CNS are predominantly found in the striatal system, as shown by in situ hybridization [30] and receptor autoradiography

* Corresponding author. Fax: (351) (1) 443-1631.

0006-8993/94/$07.00 © 1994 ELsevier Science B.V. All rights reserved SSDI 0006-8993(94)00347-F

[13,26], using the highly selective A2a agonist, COS 21680 [12].

However, some studies showed the existence of [3H]CGS 21680 binding sites also in the hippocampus and cortex [11,15,33]. Although many studies have demonstrated exclusively inhibitory effects of adeno- sine derivatives in the hippocampus [8,17], some stud- ies showed that low nanomolar concentrations of CGS 21680 [3,31] or adenosine [23,24] enhance neuronal excitability and neurotransmitter release in the hip- pocampus.

In the present study, we investigated whether adenosine A2a receptor m R N A is expressed in the CA1, CA3 and dentate gyrus areas of the hippocam- pus, using thermocycling analysis and in situ hybridiza- tion. The distribution of the adenosine A2a receptor in

R.A. Cunha et al. / Brain Research 649 (1994) 208-216 209

the hippocampus was investigated by receptor autora- diography, using the A2a-selective ligand, [3H]CGS 21680. Finally, we investigated whether stimulation of A2a receptors with the selective agonist CGS 21680 affected the ability of the A I agonist CPA to inhibit neuronal excitability.

2. Materials and methods

2.1. Thermocycling analysis"

For thermocycling analysis, male Sprague-Dawley rats (180-200 g) were anesthetized with CO 2 and decapitated. The hippocampus and striatum of each cerebral hemisphere and the right posterior cortex were immediately dissected out and frozen on dry ice. The three parts of the hippocampus were separated according to previous description [16]. Two hippocampi of each of three rats were removed and slices were transversely cut 500 /zm thick. The eight central slices of each hippocampus were put in ice-cold PBS (130 mM NaCI and 3 mM KH2PO4, pH 7.4) and dissected under a microscope on a rubber surface over a drop of ice-cold PBS. The CA1, CA3 and dentate gyrus areas were separated and immediately frozen on dry ice. The frozen tissue was homogenized in 4 M guanidine thio- cyanate in 25 mM sodium citrate (pH 7.1) with 2% (v/v) fl- mercaptoethanol by 5 up and down strokes (1,500 rpm) in a Teflon/ glass homogenizer. Poly(A)-mRNA was directly isolated by hy- bridization with a synthetic biotinylated oligo(dT) probe captured by streptavidin paramagnetic particles in a magnetic stand (PolyAT- tract R System 1000, Promega, Scandinavian Diagnostic Services, Falkenberg, Sweden). After spectrophotometrical determination of the RNA concentration, the RNA was precipitated overnight with one-tenth volume of 3 M sodium acetate (pH 5.2) and one volume of isopropanol at -20°C and resuspended in diethyl pyrocarbonate- treated water at a final concentration of 25 Izg/ml. First-strand cDNA was synthesized by reverse transcription of 50 ng of poly(A)- mRNA with 10 U/ml Moloney murine leukemia virus reverse tran- scriptase (Gibco/BRL, Grand Island, NY, USA) in the presence of 25 ~g/ml pd(N) 6 random hexamers (Pharmacia-LKB, Uppsala, Swe- den), 16 mM dNTP (Pharmacia-LKB), 1 U / ~ I rNasin Ribonuclease Inhibitor (Promega, Sweden) in a buffer containing 5 mM MgCI 2, 50 mM KC1, 10 mM Tris-HCl (pH 9.0) and 0.1% (v/v) Triton X-100 in a final volume of 20/~1. The samples were preincubated at 23°C for 10 min and the reaction was continued for 40 min at 42°C. The reaction was stopped by heating for 5 min at 99°C. Oligonucleotide primers, containing an EcoRI linker, for human [9] or rat [7] A2a receptors, corresponding to amino acids NLQNVT (5'-CGAA- TTCAACCTGCAGAACGTCACC-3', sense) and to amino acids IAIDR (5'-TCGAATTCGCGGGTC(G/ A )ATGGCGAT(A/ G)-3', antisense) were synthesized (Scandinavian Gene Synthesis, K6ping, Sweden). The sequence between the A2a receptor primers corre- sponds to a 67 amino acid sequence between the first and the beginning of the second cytoplasmic loop of the receptor. Second- strand synthesis of 10 ~1 cDNA and subsequent thermocycling with 0.025 U//~I of Thermoactinomyces aquaticus polymerase (Promega, Sweden) were performed at a primer concentration of 200 nM, in a buffer containing 10 mM Tris-HC1 (pH 9.0), 50 mM KCI, 0.1% (v/v) Triton X-100, 4 mM MgCI 2 in a volume of 50 /xl. After 1 min at 94°C, 35 cycles of 1 min at 92°C (denaturation), 2 rain at 55°C (annealing) and 2 min at 72°C (extension) were carried out, with a final extension for 5 min at 72°C, in a volume of 50 ~1 in a MiniCycler (MJ Research, Cambridge, MA, USA). Products were analyzed by electrophoresis in 2% (w/v) ultrapure agarose (BRL, Life Technologies, Gaithersburg, MD, USA).

2.2. In situ hybridization

In situ hybridization studies to detect adenosine A2a and A 1 receptor mRNA were performed as previously described [14]. Briefly, male Sprague-Dawley rats (180-200 g) were killed by decapitation after CO 2 anesthesia. The brains were dissected out and frozen in dry ice. Coronal sections, 14 ~m thick, were cut with a Leitz microtome-cryostat, thaw-mounted on poly-L-lysine coated slides and stored at -20°C. A 44-mer oligonucleotide probe for the A2a recep- tors, complementary to nucleotides 916-959, encoding an amino acid sequence between the 5th and 6th transmembrane segments of the dog A2a receptor [29], and a 48-mer oligonucleotide probe for the A~ receptors, complementary to nucleotides 985-1032 of the rat A~ receptor [18] were synthesized (Scandinavian Gene Synthesis). The A2a probe (5'-CGAGCCCGCTCCCCAGGCAGAGGCTGGCTC- TCCATCTGCTTCAG-3') was designed to match a characteristic rat A2a cDNA sequence [7] and to minimize homology with the rat A l gene sequence; overall there are 22 matches with the rat A t cDNA [18], but with two sequence gaps. The probes (80 ng) were radiola- belled by incubation at 37°C for 90 min with 1 U/ml of terminal transferase (Pharmacia-LKB) and 2 mCi/ml of [a-35S]dATP (NEN Research Products, DuPont) in One-Phor-All Plus buffer (Pharmacia) in a final volume of 25 p.l. The sections were hybridized (15 h at 42°C) with 107 cpm/ml (approximately 64 ng/ml) of the radiolabelled probes in a solution containing 50% deionized for- mamide (Fluka), 4×SSC (I×SSC is 0.15 M NaCI and 0.015 M sodium citrate, pH 7.0), 1 × Denhardt's solution (Sigma), 1% sarcosyl (Sigma), 0.02 M NaPO 4 (pH 7.0), 10% dextran sulfate (Sigma), 0.5 mg/ml of yeast tRNA (Sigma), 0.1 mg/ml of sheared salmon sperm DNA (Sigma) and 0.06 M dithiothreitol. Non-specific hybridization was assessed in the presence of 6 p.g/ml of cold probe. After hybridization, the sections were washed in 150 mM NaCI and 15 mM sodium citrate (pH 7.4) (4× 15 min at 45°C for the A2a probe or at 55°C for the A 1 probe), dipped in water and in increasing concentra- tions of ethanol, and dried. The sections were apposed to Hyperfilm- /3max (Amersham) for a period of 12-14 weeks (A2a) or 4 weeks (A1), and the autoradiograms obtained (X-Ray developer LX-24 and Unifix, Kodak) analyzed with a Microcomputer Imaging device (Imaging Research Inc., Canada).

2.3. Receptor autoradiography

Autoradiographic studies to detect adenosine A2a and A 1 recep- tors were performed as previously described [14], except that the concentration of [3H]CGS 21680 used (30 nM) was within the concentration range that facilitates neuronal activity in the hip- pocampus [31]. Briefly, Sprague-Dawley rats (180-200 g) were decap- itated, the brain dissected out and frozen on dry ice. Coronal sections (10/zm thick) were cut with a Leitz microtome-cryostat and thaw-mounted on gelatin-coated slides. The sections were pre-in- cubated for 30 min at 37°C in 170 mM Tris-HC1 buffer (pH 7.4) containing 1 mM EDTA and 2 U/ml adenosine deaminase (calf intestine; Boehringer Mannheim) and then washed twice for 10 min at 23°C in 170 mM Tris-HCl buffer with 10 mM MgCl 2. Incubations were performed for 2 h at 23°C in 170 mM Tris-HC1 (pH 7.4) containing 10 mM MgCl~, 2 U /ml adenosine deaminase and 30 nM [3H]2-[ p-(2-carboxyethyl)phenylethylamino]-5'-N-et hylcarboxamido- adenosine (CGS 21680; New England Nuclear) or 1 nM [3H]N6- cyclohexyladenosine (CHA; New England Nuclear). Non-specific binding was determined in the presence of 100 /xM 2-chloroadeno- sine (Sigma; for [3H]CGS 21680) or 30 #M R-PIA (Boehringer Mannheim; for [3H]CHA). The sections were then washed twice for 5 min in ice-cold Tris-HCl buffer (170 mM, pH 7.4) containing 10 mM MgCI2, dipped quickly three times in ice-cold distilled water and dried at 4°C over a strong fan. The dried sections were then

210 R.A. Cunha et al . /Brain Research 649 (1994) 208 216

apposed to Hyperfilm-3H (Amersham) for 12-14 weeks (A 2~,) or 3 weeks (A t) and the resulting autoradiograms were analyzed with a microcomputer imaging device (Imaging Research). Optical density was converted to density of bound ligand with the help of co-exposed standards (Amersham) containing known amounts of tritium.

3. Results

3.1. Expression of A 2a receptor mRNA

2.4. Electrophysiological recordings

The experiments were performed on hippocampal slices from male Wistar rats (181)-200 g). The animals were decapitated and one hippocampus dissected free in ice-cold Krebs solution of the follow- ing composition (mM): NaC1 124, KC1 4, MgC12 2, CaCI 2 2, glucose 10, NaH2PO 4 1.25, NaHCO 3 25, gassed with 95% 0 2 / 5 % CO z. Slices (400 p,m thick) were cut perpendicular to the long axis of the hippocampus and allowed to recover for at least 1 h in gassed Krebs solution, at room temperature. A slice was then transferred to a 1 ml recording chamber for submerged slices and continuously superfused with gassed Krebs solution kept at 30.5°C, at a flow rate of 4 ml /min . Drugs were added to this superfusion solution.

Monopolar stimulation (rectangular pulses of 0.1 ms applied with a frequency of 0.1 Hz) was delivered through a bipolar concentric electrode placed on the Schaffer col la teral /commissural fibers, in the stratum radiatum near the C A 3 / C A 1 border. Orthodromically- evoked population spikes were recorded through an extracellular microelectrode (4 M NaC1, 2 -4 Mg/ resistance) placed in the stra- tum pyramidale of the CAI area. Occasionally, evoked field excita- tory postsynaptic potentials (EPSPs) were recorded with a similar extracellular microelectrode placed in the stratum radiatum of the CAI area. When recording only population spikes, the intensity of the stimulus (200-400 mA) was adjusted to evoke a response with an amplitude near 50% of the population spike amplitude obtained with supramaximal stimulation. When recording EPSPs, the intensity of the stimulus (190-290 mA) was adjusted to obtain a large EPSP slope with a minimum population spike contamination. The individ- ual responses were displayed on a Tektronix (2430A) digitizing oscilloscope and the averages of 8 o1 16 consecutive responses were digitally recorded on a Tekmate computer. Responses were quanti- fied either as the amplitude of the averaged population spikes or as the slope of the initial phase of the averaged EPSPs.

The data are expressed as mean+S .E .M, from n observations, and the significance of the means was calculated by the Student 's t-test. The null hypothesis was rejected at P < 0.05. CGS 21680 and N%cyclopentyladenosine (CPA) were from Research Biochemicals and were made up in 10 mM stock solutions in DMSO. Aliquots of these stock solutions were kept frozen ( - 2 0 ° C ) until use on the day of the experiment.

Fig. 1A shows the electrophoretic profile of the thermocycling analysis products obtained with 50 ng of poly(A)-mRNA from the hippocampus as template; the results obtained with poly(A)-mRNA from the cerebral cortex and striatum are also shown for comparison. It can be seen that in all three brain structures, a band was obtained with a number of base pairs similar to the expected length (216 bp) of the sequence of the cloned Aza receptor between the two chosen primers [7,9]. The thermocycling analysis signal obtained in the cere- bral cortex and hippocampus was similar to that ob- tained in the striatum, where the A 2 a receptor is well characterized [12,26]. Similar results were obtained upon identical thermocycling analysis in three other experiments (one animal per experiment). When the extracted poly(A)-mRNA was subjected to reverse t ranscr iptase/ thermocycl ing analysis in the absence of reverse transcriptase, no thermocycling analysis signal was obtained, indicating that A 2, DNA segments were not present as contaminants in the extracted poly(A)- mRNA. Within the hippocampus, the A 2 a receptor m R N A was expressed in the CAI, CA3 and dentate gyrus, as shown in Fig. lB. Similar results were ob- tained in two other experiments.

The expression of the A 2,~ m R N A in the three areas of rat hippocampus was further confirmed by in situ hybridization (Fig. 2A, C). Identical experiments per- formed in three different animals gave similar results. The A 2~ m R N A was mainly found in the region of the cell bodies of the granular cell layer of the dentate gyrus and at the stratum pyramidale of the cornu Ammonis (Fig. 2A). This anatomical distribution closely matched that of the A 1 m R N A expression (Fig. 2B), although the relative proportions of A 2:, and A t mRNA

St S H C - St 1 3 D G -

Fig. 1. Expression of Aza m R N A in the striatum, cortex and hippocampus of the rat and in the CA1, CA3 and dentate gyrus areas of the hippocampus. Poly(A)mRNA was extracted from these brain areas, reverse transcribed into cDNA, subjected to thermocycling analysis and the products analyzed by agarose gel electrophoresis. The number of base pairs of the thermocycling products was assessed by co-analysis with eDNA standards (123 bp ladder); St, standards; S, striatum; H, hippocampus; C, cortex; - , control (i.e. no poly(A)-mRNA added during the termocycling analysis); 1, CA1; 3, CA3; DG, dentate gyrus. The gel electrophoresis analysis of the three brain regions is representative of four identical experiments (one animal per experiment) and the analysis of the three areas of the hippocampus is representative of three identical experiments (three animals per experiment).

R.A. Cunha et al. / Brain Research 649 (1994) 208-216 211

A C

B D

E G

F H

/

i"

Fig. 2. Anatomical distribution of the A2a and A 1 mRNA expression and of the binding sites of A2a agonist [3H]CGS 21680 and of the A l agonist [3H]CHA in the rat hippocampus. A and B show in situ hybridization autoradiograms obtained with a probe for the adenosine A2a receptor mRNA (A) and for the A 1 mRNA (B), in a 14/xm coronal section of the brain, restricted to the hippocampal formation. C and D show the corresponding non-specific labelling obtained in adjacent sections, in the presence of a 1000-fold higher concentration of non-labelled A2a mRNA probe (C) or of non-labelled A 1 mRNA probe (D). E and F show autoradiograms obtained upon incubation with 30 nM [3H]CGS 21680 (E) or 1 nM [3H]CHA in a 10 mm coronal section of a rat brain, restricted to the hippocampal formation. G and H show the corresponding non-specific labelling obtained in adjacent section, in the presence of 100 ,~M 2-chloroadenosine (G) or 30 ,o,M R-PIA (H).

212 R,A. Cunha et al. / Brain Research 049 (1994) 207¢ 21h

expressed in each area seemed to differ: the optical densities in the autoradiograms for the A2, mRNA in situ hybridizations over the pyramidal cell layer of CA3, the granule cell layer of the dentate gyrus and the pyramidal cell layer of the CA1 were related to each other as 0.8:0.8:1; the corresponding ratios for the A I mRNA were 1.9:1.1:1.0 (see also [14]).

3.2. Identification of [ 3H]CGS 21680 binding sites

Fig. 2E shows receptor autoradiograms obtained in one out of three similar experiments where [3H]CGS 21680 (30 nM) was used and the exposure time was 13 weeks. Under these conditions there was an intense binding of [3H]CGS 21680 to the hippocampal forma- tion (Fig. 2E). The number of binding sites was about 25% of the [3H]CGS 21680 binding in the anterior caudate-putamen which was 332 _+ 8 fmol /mg of grey matter. Specific binding of [3H]CGS 21680 in the hip- pocampus was equivalent to more than 90% of the total binding, since non-specific binding was close to background (Fig. 2G). The most densely labelled area was the stratum radiatum of the CA1 area. The distri- bution of [3H]CGS 21680 binding within the hippocam- pus is shown in Table 1 and matched the distribution of the [3H]CHA binding sites in the hippocampus (Fig. 2F). The main difference was that [3H]CGS 21680 binding was greater than [3H]CHA binding in the stratum lacunosum of the CA3 area and lower in the subiculum (cf. [6]).

/ " 1OhM

J

g ins

C G S 2 1 6 8 0

10nM

Fig. 3. Effect of the selective adenosine A2a receptor agonist, CGS 21680 (10 nM), on orthodromically evoked population spikes (PS) (upper part) and excitatory postsynaptic potentials (EPSPs) (lower part) recorded simultaneously with extracellular electrodes placed in the CAI area of a rat hippocampal slice. Each trace is the average of 16 consecutive responses, which are preceded by the presynaptic volley and the stimulus artefact. The responses in the presence of CGS 21680 are indicated by the arrow and correspond to the full effect of CGS 21680 (10 nM). The control response was recorded immediately before starting the superfusion with CGS 21680.

3.3. Interaction between excitatory adenosine A2a and inhibitory adenosine A ~ receptor actiuation

Fig. 3 shows the effect of CGS 21680 (10 nM) on population spikes and field excitatory postsynaptic po-

Table 1 Anatomical distribution of [3H]CGS 21680 binding in the rat hip- pocampus

Specific binding ( f m o l / m g of grey matter)

CA1 oriens 77_+6 radiatum 98 ± 6 moleculare 48 ± 2

CA3 oriens 71 ± 3 radiatum 80 ± 4 moleculare 65 ± 1

Dentate gyrus supra-granular 81 ± 3

Subiculum 84 ± 6

The data were calculated from autoradiograms obtained upon incu- bation of 30 nM [3H]CGS 21680 with 10 ~ m coronal sections of a rat brain and exposed for at least 12 weeks. The data are mean ± S.E.M. of three animals.

tentials (EPSPs) recorded simultaneously from the CA1 region of a rat hippocampal slice upon stimulation of the Schaffer collateral/commissural afferents. It can be observed that CGS 21680 (10 nM) increased both the amplitude of the population spikes and the slope of the EPSPs. The average increase in population spike amplitude caused by CGS 21680 (10 nM) was 17 +_ 2% in six slices from different animals, where the ampli- tude of the population spikes before applying CGS 21680 was 1.82 + 0.33 mV. The average increase in the slope of the EPSPs caused by CGS 21680 (10 nM) was 32 + 5% in three experiments, where the EPSP slope before applying CGS 21680 was 0.40_+ 0.04 mV/ms. The selective adenosine A l receptor agonist, CPA (6-15 nM) caused a concentration-dependent depres- sion of the amplitude of orthodromically-evoked popu- lation spikes (Fig. 4). The average inhibition of popula- tion spike amplitude caused by CPA (6 nM) was 39 + 3% in six slices in which the average control population spike amplitude was 2.8 _+ 0.3 mV.

To determine if activation of adenosine A2a recep- tors in the hippocampus [31] could influence the ability of the adenosine A I receptors to inhibit hippocampal excitability, the effect of the selective A l agonist, CPA (6 nM) [22] on population spike amplitude was studied

R.A. Cunha et al. / Brain Research 649 (1994) 208-216 213

in the absence and in the presence of the selective A2a agonist, CGS 21680 (10 nM) [12]. The effect of CPA (6

nM) was tested first in the absence of CGS 21680; CP A was then washed out and CGS 21680 (10 nM) was applied. Af te r a stable response to CGS 21680 had been ob ta ined (within 20 to 35 min), the effect of a second appl icat ion of C P A (6 nM), now in the pres- ence of CGS 21680 (10 nM), was tested. Fig. 5 shows that the inhibi t ion of popula t ion spike ampl i tude caused by CPA (6 nM) was a t t enua ted in the presence of CGS 21680 (10 nM). The average decrease in popu- lat ion spike ampl i tude caused by CPA (6 nM) was 32 + 5% in the absence of CGS 21680 and only 13 + 7% ( P < 0.05 as compared with the control effect of CPA)

A

.?.=~

1OO

75

50

CPA (6nM) j~ I

' 0 ~

°~oO • 0

• 000000

t~ o 0 0 0 0 - ~ o , . • • •

0

0 e e l

Oo

L J 0 50 1 0 0

Time (min) AI" "0 . m

E <

1

12.

C P A (nM) c 1'o ~'~

°'I%°°°

"t.1.- %;" .,..

° o

I, 0 3~5 7JO

T ime (rain)

W

0 0

I 140

l m V

C 6 10 15 W

lOres

Fig. 4. Effect of the selective adenosine A 1 receptor agonist CPA on the average amplitude of orthodromically-evoked population spikes (PS) recorded extracellularly from the CA1 pyramidal cell layer of a rat hippocampal slice. The upper part represents the time-course of the effect of CPA. Each point corresponds, in the ordinates, to the average amplitude of 16 successive population spikes, and in the abscissae, to the start of averaging. The beginning of the superfusion of each concentration of CPA and of the washout (w) of CPA is indicated by the arrows; c represents the control population spike amplitude. The lower part shows the pen-recorder traces of the average population spikes corresponding to 8 min (c, control), 38 min (6 nM CPA), 65 min (10 nM CPA), 92 min (15 nM CPA) and 135 rain (w, washout) in the upper part; each population spike is pre- ceded by the presynaptic volley and the stimulus artefact.

B CPA (6nM)

50

' < D ~ a e

o CGS 21680 (lOnM) - ,

1.0-

E o.5-

o-

Fig. 5. Attenuation by the selective adenosine A2a receptor agonist, CGS 21680, of the inhibitory effect of the selective adenosine A 1 receptor agonist, CPA, on the average amplitude of orthodromically- evoked population spikes (PS) recorded extracellularly from the CA1 pyramidal cell layer of rat hippocampal slices. A: Time course of the effect of CPA (6 nM) in the absence (filled circles) and in the presence (open circles) of CGS 21680 (10 nM), in the same hip- pocampal slice. The ordinates represent the population spike ampli- tude as percentage of its value before applying CPA. In control conditions (filled circles) 100% corresponds to 2.8 mV; in this experiment CGS 21680 (10 nM) increased population spike ampli- tude by 17%. The abscissae represent the start of averaging. The horizontal bar represents the period of CPA (6 nM) superfusion. B: Pooled data from three experiments similar to those illustrated in A. The ordinates represent the decrease in population spike amplitude as a percentage (left panel) or as absolute values (AmV) (right panel). In the left panel, 0% is the averaged population spike amplitude before applying CPA and 100% represents a complete inhibition of population spike amplitude; in the absence of CGS 21680, 100% corresponds to 2.4+0.3 mV. In these experiments CGS 21680 (10 nM) increased population spike amplitude by 15:1: 1%. The absence ( - ) or presence (+) of CGS 21680 is indicated below each column. * P < 0.05 (one-tailed paired Student's t-test) as compared with the effect of CPA in the absence of CGS 21680.

in the presence of CGS 21680 (10 nM) (Fig. 5B). The effect of C P A (6 nM), expressed as the absolute de- crease in popula t ion spike ampl i tude (zamV, Fig. 5B),

was also smaller ( P < 0.05) in the presence of CGS 21680 (10 riM). In one exper iment , the intensi ty of

214 R.A. Cunha et al. / Brain Research (~4q ( lq04) 208-216

orthodromic stimulation after applying CGS 21680 was adjusted so that the amplitude of the population spike before applying CPA, in the absence and in the pres- ence of CGS 21680, was similar (1.8 mV). Under these conditions it was also observed that the decrease in population spike amplitude caused by CPA (6 nM) was smaller (0% decrease) in the presence of CGS 21680 (10 nM) than in its absence (27% decrease). CPA (15 nM) decreased population spike amplitude by 68% in the absence of CGS 21680 and by only 33% in the presence of CGS 21680 (10 nM). Thus, the degree of attenuation of the effect of CPA (15 nM) by CGS 21680 (10 nM) was quantitatively greater than the excitatory effect of CGS 21680 (10 nM) alone. This further precludes the possibility that the ability of CGS 21680 to attenuate the effect of CPA is solely due to the increase in population spike caused by CGS 21680. Two consecutive applications of CPA (6 nM) in control conditions caused a similar inhibition of population spike amplitude, i.e., the ratio between the effect of the second CPA application over the effect of the first CPA application was 0.97 + 0.14 (n = 2).

4. Discussion

The major findings of the present study were (1) that A2a mRNA is expressed in the rat hippocampus, (2) that [3H]CGS 21680, in a concentration where it is reported to selectively bind to adenosine A2~ receptors [13], can label adenosine receptors in rat hippocampal slices, and (3) that the ability of the adenosine A I receptor agonist, CPA [22] to inhibit population spike amplitude was attenuated by the A2a agonist CGS 21680, in a low nanomolar concentration.

Thermocycling analysis and in situ hybridization ex- periments show the expression of mRNA for the adenosine A2~ receptor in the CA1, CA3 and dentate gyrus areas of the hippocampus. The distribution of the A2a mRNA expression found in the present work resembled, but was not identical to, that of the A~ mRNA expression [14], suggesting that there might be a co-expression of A2~ and A 1 mRNA. This idea was strengthened by the observation that the distribution of the binding sites of the A2a agonist, [3H]CGS 21680, was similar to the previously described [6] distribution of the binding sites of the A 1 agonist [3H]CHA.

Previous autoradiographic studies to label CGS 21680-sensitive adenosine A2a receptors in the brain concluded that these receptors are exclusively located in the basal ganglia [13,19]. However, in these studies, a 6-times lower concentration of [3H]CGS 21680 than in the present study was used and the time of exposure was at least 4-times shorter. It is possible that in the hippocampus the density of adenosine A2a receptors is lower, a n d / o r their affinity for CGS 21680 is smaller

than in the basal ganglia. This might explain why labelling those receptors requires a slightly higher [3H]CGS 21680 concentration and a longer exposure time in the hippocampus than in the basal ganglia. It was recently shown [15] that the binding of [3H]CGS 21680 to the rat cerebral cortex differs from its binding to the striatum. The hypothesis that CGS 21680 might be interacting with a cortical adenosine receptor differ- ent from the A2a receptor could not be ruled out [15]. Thus, the possibility that the [3H]CGS 21680 binding found in the hippocampus in the present study may at least partly represent binding to another receptor than the Aza receptor cannot be ruled out. Despite the overall similarity in the distribution of CGS 21680 and CHA binding sites, the [3H]CGS 21680 binding sites are unlikely to be A1 receptors, since CGS 21680 has a very low affinity for A~ receptors [10,12]. We favor the view that at least a proportion of these receptors in the hippocampus are A 2~, receptors. A low density of bind- ing sites for [~H]CGS 21680 in homogenates of hip- pocampal membranes, with the high affinity expected for an A2a receptor, have previously been described [11,33]. We have recently confirmed the presence of such high affinity binding sites (Johansson, Cunha, Sebastifio and Fredholm, in preparation), even though there appear to be sites with a lower affinity, uncharac- teristic of archetypical A ,~, receptors.

The presence of a low density of A~,, receptors to which could bind [3H]CGS 21680 in the hippocampus is further suggested by a low, but distinct A)~, mRNA expression in this structure. The results obtained with the in situ hybridization experiments to detect mRNA for the adenosine A2a receptor showed that a signal clearly above the background could be observed when an exposure time of at least 12 weeks was used. That previous in situ hybridization studies [7,14,30] have not detected A2, , mRNA in the hippocampus, might be due to the short exposure time used in these studies. However, the long exposure with the A a, mRNA probe used in this study produced a substantial non-specific signal in the granular and pyramidal cells of the hip- pocampus. Similar non-specific hybridization with long exposure periods has been described to occur in the pyramidal and granular cells of the hippocampus with other probes [28]. It is unlikely that the in situ hy- bridization signal obtained with the Aa, probe is due only to hybridization with the more abundant A~ mRNA, since thermocycling results also showed an A ,~, mRNA expression in the hippocampus. The somewhat different density of distribution of the A~ and A2a mRNAs in the hippocampus also argues against this possibility, as does the rather low degree of sequence homology in the area used as template for the probe.

The presence of adenosine Aa, receptors in the hippocampus can also be inferred from the observa- tions that low nanomolar concentrations of CGS 21680

R.A. Cunha et al. / Brain Research 649 (1994) 208-216 215

enhance [3H]acetylcholine release from the CA3 and dentate gyrus areas [3] and enhance neuronal excitabil- ity in the CA1 area [31]. Nanomolar concentrations of adenosine also enhance neuronal excitability in the three areas of the hippocampus [23,24]; this may result from activation of CGS 21680-sensitive adenosine A2a receptors. Some effects of high concentrations of ade- nine derivatives, on P-type Ca 2÷ currents [21] and on cAMP accumulation [17], are not mimicked by CGS 21680 and are probably due to activation of adenosine A2b receptors.

The only regions consistently labelled with the A2a receptor mRNA probe were the stratum pyramidale of the cornu Ammonis and the stratum granulare of the dentate gyrus. Within each area of the cornu Ammo- nis, the greatest density of [3H]CGS 21680 binding was found in the stratum radiatum, where the synaptic connections between granular and pyramidal cells or between pyramidal cells are located. This suggests a possible, probably excitatory, role in synaptic transmis- sion that is supported by the finding that CGS 21680, at a low nanomolar concentration, increased the slope of the extracellularly recorded EPSPs, as well as in- creasing the amplitude of population spikes recorded from the CA1 pyramidal cells. The granular and pyra- midal cells may be sites of A 2a mRNA expression, and the A 2a receptor may be located either in the dendrites or in the terminals of the granular and pyramidal cells. A pre-synaptic co-localization of excitatory adenosine A2a and inhibitory adenosine A 1 receptors has been demonstrated at a rat neuromuscular junction [2]. The distribution of A2a mRNA and of [3H]CGS 21680 binding in the hippocampus does not exclude that some GABAergic interneurons located in the vicinity of the pyramidal and granular cells might also express A2a mRNA. The present observation that A2a recep- tor mRNA is expressed in the cerebral cortex consti- tutes evidence for the presence of adenosine A2a re- ceptors in this brain area, where CGS 21680 inhibits the ischemia-evoked release of GABA [25].

There appeared to be at least a gross co-expression and co-localization of the A2a and A~ receptors in the hippocampus. The functional data indicated an inter- action between these receptors, providing a mixture of excitatory and inhibitory modulation by adenosine on neuronal excitability in one hippocampal pathway. In- deed, it was observed that, in the presence of a low nanomolar concentration of CGS 21680, the ability of CPA to inhibit population spike amplitude was attenu- ated. This attenuation was observed both when the inhibition of the effect of CPA was quantified in terms of percentage decrease and as absolute decrease in population spike amplitude. The attenuation of the effect of CPA by CGS 21680 was not dependent on the population spike amplitude. An excitatory effect of nanomolar concentrations of CGS 21680 in hippocam-

pal excitability, probably due to activation of adenosine A2a receptors, was recently reported [31]. The in- hibitory effects of nanomolar concentrations of CPA are well known to be due to activation of adenosine A~ receptors [22], which are present in high amounts in the hippocampus [6]. The inhibitory adenosine recep- tors in the CA1 region are A~ receptors [5,32]. It thus appears that activation of adenosine A2a receptors in the hippocampus reduces the ability of adenosine A~ receptors to inhibit hippocampal excitability. This sug- gests that there is a cross-talk between A~-inhibitory and A2a-excitatory adenosine receptors in the hip- pocampus, and that the net modulation of hippocam- pal excitability by adenosine derivatives depends on a balance between the degree of activation of both adenosine receptors.

The recent reports that low nanomolar concentra- tions of CGS 21680 may enhance acetylcholine release [2,3,27] or modulate the release of GABA [20,25] sug- gest that CGS 21680-sensitive adenosine A2a receptors may interact with several transducing systems in differ- ent preparations. The precise nature of these transduc- ing systems and how they interact remains to be inves- tigated.

In summary, we have observed that CGS 21680-sen- sitive adenosine A2a receptors in the rat hippocampus can be labelled in situ and that A2a receptor mRNA is expressed in the CA1, CA3 and dentate gyrus areas of the hippocampus, as well as in the cerebral cortex. It was also observed that activation of adenosine A2a receptors in the hippocampus may decrease the ability of adenosine A~ receptors to modulate hippocampal activity. These observations suggest that the A2a-medi- ated neuromodulation is not confined to the basal ganglia, but may be a more widespread neuromodula- tory phenomenon in the central nervous system.

Acknowledgements

R.A.C. was in receipt of a European Neuroscience Programme Short-Term Fellowship and was supported by JNICT (BD/148-90/ID). B.J. is supported by a Swedish Medical Research Council fellowship. This work was partly supported by a Swedish Medical Re- search Council grant (proj. no. 2553), the Swedish Society for Medical Research, a JNICT grant (STRDA/ C / SAU/ 292/ 92) and grants from Karolin- ska Institutet and Astra Arcus AB. The authors are involved in a concerted action ADEURO (EEC, Biomed 1).

References

[1] Brown, S.J., James, S., Reddington, M. and Richardson, P.J., Both A 1 and A2a purine receptors regulate striatal acetyl- choline release, J. Neurochem, 55 (1990) 31-38.

216 R.A. Cunha et al. /Brain Research ()4g (lqg4l 208 216

[2] Correia-de-Sfi, P., SebastiS.o, A.M. and Ribeiro, J.A.. Inhibitory and excitatory effects of adenosine receptor agonists on evoked transmitter release from phrenic nerve endings of thc rat, Br. Z Pharmacol., 103 (1991) 1614-162/).

[31 Cunha, R.A., Milusheva, E., Vizi, E.S., Ribeiro, J.A. and Sebas- ti~o, A.M., Exitatory and inhibitory effects of A~ and A2,x adenosine receptor activation on the electrically-evoked [3H]- acetylcholine release from different areas of the rat hippocam- pus, J. Neurochem., 63 (1994) in press.

[4] Dunwiddie, T.V., The physiological role of adenosine in the central nervous system, Int. Rev. Neurobiol., 27 (1985) 63-139.

[5] Dunwiddie, T.V. and Fredholm, B.B., Adenosine receptors me- dialing inhibitory electrophysiological responses in rat hip- pocampus are different from receptors mediating cyclic AMP accumulation, Naunyn-Schmiedeberg's Arch. Pharmacol., 326 (1984) 294-301.

[6] Fastbom, J., Pazos, A. and Palacios, J.M., The distribution of adenosine A I receptors and 5'-nucleotidase in the brain of some commonly used experimental animals, Neuroscience, 22 (1987) 813 826.

[7] Fink, J.S., Weaver, D.R., Rivkees, S.A., Peterfreund, R.A., Pollack, A.E., Adler, E.M. and Reppert, S.M., Molecular cloning of the rat A 2 adenosine receptor: selective co-expression with D 2 dopamine receptors in rat striatum, Mol. Brain Res., 14 11992) 186-195.

[8] Fredholm, B.B., Adenosine At-receptor-mediated inhibition of evoked acetylcholine release in the rat hippocampus does not depend on protein kinase C, Acta Physiol. Scand., 140 (1991)) 245 -255.

[9] Furlong, T.J., Pierce, K.D., Selbie, L.A. and Shine, J., Molecular characterization of a human brain adenosine A 2 receptor, Mol. Brain Res., 15 11992) 62-66.

[10] Hutchison, A.J., Webb, R.L, Oei, H.H., Ghai, G.R., Zimmer- man, M.B. and Williams, M., CGS 21680C, an A 2 adenosine receptor agonist with preferential hypotensive activity, J. Phar- macol. Exp. Ther., 251 11989) 47-55.

[11] James, S., Xuereb, J.H., Askalan, R. and Richardson, P.J., Adenosine receptors in post-mortem human brain, Br. J. Phar- macol., 105 11992) 238-244.

[12] Jarvis, M.F., Schulz, R., Hutchison, A.J., Do, U.H., Sills, M.A. and Williams, M., [3H]CGS 21680, a selective A , adenosine receptor agonist, directly labels A 2 receptors in rat brain, J. Pharmacol. Exp. Ther., 251 (1989)888-893.

[13] Jarvis, M.F. and Williams, M., Direct autoradiographic localiza- tion of adenosine A 2 receptors in the rat brain using the A2-selective agonist, [3H]CGS 21680. Eur. J. Pharrnacol., 168 (1989) 243-246.

[14] Johansson, B., Ahlberg, S., van der Ploeg, 1., Bren6, S., Linde- fors, N., Persson, H. and Fredholm, B.B., Effect of long term caffeine treatment on A~ and A 2 adenosine receptor binding and on mRNA levels in rat brain, Naunyn-Schmiedeberg's Arch. Pharmacol., 347 (1993) 407-414.

[15] Johansson, B., Georgiev, V., Parkinson, F.E. and Fredholm, B.B., The binding of the adenosine A 2 receptor selective ago- nist [3H]CGS 21680 to rat cortex differs from its binding to rat striatum, Eur. J. Pharmacol. Mol. Pharmacol. Sect., 247 (1993) 1 0 3 -11 (I.

[16] Lopes da Silva, F.H. and Arnold, D.E.A.T., Physiology of the hippocampus and related structures, Annu. Rec. Physiol., 40 (1978) 185-216.

[17] Lupica, C.R., Cass, W.A., Zahniser, N.R. and Dunwiddie, T.V., Effects of the selective adenosine A 2 receptor agonist CGS 21680 on in vitro electrophysiology, cAMP formation and

dopamine release in rat hippocampus and slriatum, J. Pharma- col. Exit. Ther., 252 (1990) 1134-1141.

[18] Mahan, L.C., McVittie, L.D., Smyk-Randall, E.M., Nakata, II., Monsma Jr., F.J., Gert;en. ( ' .R. and Sibley, I).R., Cloning and expression of an A I adenosine receptor l'r¢ml rat brain, Mot. Pharmacol., 41111991 ) 1~ 7.

[19] Martinez-Mir, M.I.. Probsl. A. and Palacios, J.M., Adenosine A a receptors: selective localization in thc human basal ganglia and alterations with disease, Neuroscience. 42 (1991) 697 71/6.

[201 Mayfield. R.D.. Suzuki. F. and Zahniser, N.R., Adenosine A,~, receptor modulation of electrically evoked endogenous GABA release from slices of rat globus pallidus, J. Neurochem., 60 (1993) 2334 2337.

[21] Mogul, D.J., Adams, M.E. and Fox, A.P., Differential actiwLtion of adenosine receptors decreases N-type but potentiates P-type Ca=~ current in hippocampal CA3 neurons. Neuron, 1011993) 327 334.

[22] Moos, W.H.. Szotek D.S. and Bruns, R.F.. N%Cycloalkyla ~ denosines. Potent A i-selective adenosine agonists. J. Med. ('hem., 28 11985) 1383 1384.

[23] Nishimura, S., Okada, Y. and Amatsu, M., Post-inhibitory exci- lation of adenosine on neurotransmission in guinea-pig hip- pocampal slices, Neurosci. Lett., 139 11992) 126-129.

[24] Okada, Y., Sakurai, T. and Mort, M., ExeitatoD' effect ol adenosine on neurotransmission is due to increase of transmit- ter release in the hippocampal slices, Neurosci. [~ett., 142 (1992) 233-236.

[25] () 'Reagan, M.It.. Simpson, R.E., Perkins. [..M. and Phillis. J.W., Adenosine receptor agonists inhibit the release of ),- aminobutyric acid (GABA) from the ischemic rat cerebral cor- tex, Brain Res., 582 11992) 22-26.

[26] Parkinson, F.E. and Fredholm, B.B., Autoradiographic evidence for G-protein coupled A s e c e p t o r s in rat ncostriatum using [~ H]CGS 21680 as a ligand, Naunyn-Sehmiedebe~:~'s Arch. Phar- macol., 342 (1990) 85-89.

[27] Richardson, P.J. and Cunha, R.A., Extracellular metabolism and function of purines in immunoisolated cholinergic nerve terminals, Neurochem. htt., 21 (Suppl.)(1992)QI0.

[28] Schalling, M., H6kfelt, T., Wallace, B., Goldstein, M., Filer, D., Yamin, C. and Schlesinger, D.H., Tyrosine 3-hyroxylase in rat brain and adrenal medulla: hybridization histochemistry and immunohistochemistry combined with retrograde tracing, Proc. Natl. Acad. Sci. USA, 83 (1986) 6208-6212.

[29] Schiffmann, S.N., Libert, F., Vassart, G., Dumont, ,I.E. and Vanderhaeghen, J.-J., A ckmed G protein-coupled protein with a distribution restricted to striatal medium-sized neurons. Possi ble relationship with D I dopamine receptor, Brain Res'., 519 (1990) 333-337.

[30] Schiffmann. S.N., Libert, F., Vassart, G. and Vanderhaeghen, J.-J., Distribution of adenosine A 2 receptor mRNA in the human brain, Neurosci. Lett., 130 11991) 177 181.

[31] Sebastifio, A.M. and Ribeiro, J.A., Evidence for the presence of excitatory A~ adenosine receptors in the rat hippocampus, Neurosci. Lett., 138 (1992) 41-44.

[32] Sebastifio, A.M., Stone, T.W. and Ribeiro, J.A., The inhibitory adenosine receptor at the neuromuscular junction and hip- pocampus of the rat: antagonism by 1,3,8-substituted xanthines, Br. J. Pharmacol., 101 (1992) 453-459.

[33] Wan, W., Sutherland, G.R. and Geiger, J.I).. Binding of the adenosine A , receptor ligand [31t]CGS 21680 to human and rat brain: evidence for multiple affinity sites. J. Neurochem., 55 (1991)) 1763-1771.